Tracker algorithm based on upload capacity in BitTorrent network

BitTorrent has proved to be the most effective and popular P2P file distribution application. In order to improve the topological features and take full advantage of each peer´s upload capacity in BitTorrent overlay, the disadvantages of topological features and the utilization of each peer´s upload capacity in BitTorrent overlay was analyzed and UC-track, a tracker algorithm based on upload capacity, was proposed. UC-track algorithm sends neighbor list to peer that contains certain percentage of peers with similar upload capacity. When a peer is connected with other peers, it drops the connection with a peer that differs greatly in upload capacity and accepts a new connection. Simulations show that, compared with default algorithm using random selection strategy, UC-track algorithm optimizes the topological features of the BitTorrent overlay and improves the efficiency of data transmission in BitTorrent system.